Exploration of Indian Startup funding data (Kaggle)

Importing Packages

In [1]:
import pandas as pd
from pandas import DataFrame, Series
import numpy as np
import sqlite3
import matplotlib.pyplot as plt
import seaborn as sns
from pysqldf import SQLDF; sqldf = SQLDF(globals()); q = getattr(sqldf, 'execute')
import warnings; warnings.filterwarnings('ignore')
from fast import *

%reload_ext autoreload
%autoreload 2

%matplotlib inline
sns.set(style='darkgrid')
pd.set_option('display.float_format', lambda x: '{:f}'.format(x))

PATH = '../data'

Reading data

In [3]:
display(df.head())
Date StartupName IndustryVertical SubVertical CityLocation InvestorsName InvestmentType AmountInUSD Remarks Date_Year Date_Month Date_Week Date_Day Date_Weekday_Name Date_Dayofweek Date_Dayofyear Month_Year Date_quarter rank
0 2017-08-01 TouchKin technology Predictive Care Platform Bangalore Kae Capital privateequity 1300000.000000 NaN 2017 8 31 1 Tuesday 1 213 8-2017 3 1.000000
1 2017-08-02 Ethinos technology Digital Marketing Agency Mumbai Triton Investment Advisors privateequity nan NaN 2017 8 31 2 Wednesday 2 214 8-2017 3 nan
2 2017-08-02 Leverage Edu consumer internet Online platform for Higher Education Services New Delhi Kashyap Deorah, Anand Sankeshwar, Deepak Jain,… seedfunding nan NaN 2017 8 31 2 Wednesday 2 214 8-2017 3 nan
3 2017-08-02 Zepo consumer internet DIY Ecommerce platform Mumbai Kunal Shah, LetsVenture, Anupam Mittal, Hetal … seedfunding 500000.000000 NaN 2017 8 31 2 Wednesday 2 214 8-2017 3 3.000000
4 2017-08-02 Click2Clinic consumer internet healthcare service aggregator Hyderabad Narottam Thudi, Shireesh Palle seedfunding 850000.000000 NaN 2017 8 31 2 Wednesday 2 214 8-2017 3 2.000000

Data Summary

In [4]:
describe(df)

======================================
No. of Rows.:2372
No. of Columns:19

======================================
Data Type Null Count Not Null Count Unique Count
Date datetime64[ns] 0 2372 695
StartupName object 0 2372 2001
IndustryVertical object 171 2201 728
SubVertical object 936 1436 1364
CityLocation object 179 2193 71
InvestorsName object 8 2364 1885
InvestmentType object 1 2371 4
AmountInUSD float64 847 1525 328
Remarks object 1953 419 69
Date_Year int64 0 2372 3
Date_Month int64 0 2372 12
Date_Week int64 0 2372 53
Date_Day int64 0 2372 31
Date_Weekday_Name object 0 2372 7
Date_Dayofweek int64 0 2372 7
Date_Dayofyear int64 0 2372 353
Month_Year object 0 2372 32
Date_quarter int64 0 2372 4
rank float64 847 1525 120
AmountInUSD Date_Year Date_Month Date_Week Date_Day Date_Dayofweek Date_Dayofyear Date_quarter rank
count 1525.000000 2372.000000 2372.000000 2372.000000 2372.000000 2372.000000 2372.000000 2372.000000 1525.000000
mean 12031073.099016 2015.792159 6.129427 24.984823 15.337690 2.029933 170.869730 2.381113 25.941639
std 64031175.128736 0.733754 3.375506 14.794864 8.710841 1.478779 103.041844 1.096652 15.681256
min 16000.000000 2015.000000 1.000000 1.000000 1.000000 0.000000 1.000000 1.000000 1.000000
25% 370000.000000 2015.000000 3.000000 12.000000 8.000000 1.000000 83.000000 1.000000 13.000000
50% 1070000.000000 2016.000000 6.000000 24.000000 15.000000 2.000000 166.000000 2.000000 25.000000
75% 6000000.000000 2016.000000 9.000000 37.000000 23.000000 3.000000 258.000000 3.000000 38.000000
max 1400000000.000000 2017.000000 12.000000 53.000000 31.000000 6.000000 365.000000 4.000000 70.000000

Data Exploration

Distribution of investments count across different cities

In [5]:
plt.subplots(figsize=(15,5))
p1 = sns.countplot(x = 'CityLocation',
              data = df,
              order = df['CityLocation'].value_counts().index)

plt.xticks(rotation=90);

Total amount invested over months

In [6]:
df_temp = df.groupby(['Month_Year']).agg({'AmountInUSD': 'sum'}).reset_index()

plt.subplots(figsize=(8,5))
sns.barplot(x = 'Month_Year', y='AmountInUSD', data = df_temp)
plt.xticks(rotation=90);

Count of Funding in each month

In [7]:
plt.subplots(figsize=(8,5))
sns.countplot(x = 'Month_Year', data = df)

plt.xticks(rotation=90);

When funding was declared

In [8]:
temp = df.groupby(['Date_Weekday_Name']).agg({'Date':'count'}).reset_index()
sns.barplot(x=temp.Date_Weekday_Name, y=temp.Date, color='violet')
Out[8]:
<matplotlib.axes._subplots.AxesSubplot at 0x1f642291ac8>

Distribution of funding for sum of below $1M

In [9]:
ax = histogram(df.AmountInUSD, upper=10**6, bins=30, color='skyblue')
ax.annotate('~ $500K', xy=(500000, 4.7e-6), xytext=(700000, 4e-6),
            arrowprops=dict(facecolor='black', shrink=0.05), fontsize=15);
plt.title('Histogram');

When funding was announced for big investments

In [10]:
temp = df[df['rank'] == 1].sort_values('Date')
sns.countplot(x=temp.Date_Weekday_Name);

Investment Type Distribution

In [11]:
temp = df.groupby(['InvestmentType']).agg({'AmountInUSD':'sum'}).reset_index()
temp.AmountInUSD = 100*temp.AmountInUSD/np.sum(temp.AmountInUSD)
g = sns.barplot(x=temp.InvestmentType, y=temp.AmountInUSD)
for index, row in temp.iterrows():
    g.text(row.name,row.AmountInUSD, round(row.AmountInUSD,2), color='black', ha="center")
plt.xticks(rotation=45);

Industry Vertical Distribution

In [12]:
temp = df.groupby(['IndustryVertical']).agg({'AmountInUSD':'sum'}).reset_index()\
  .sort_values('AmountInUSD', ascending=False).reset_index(drop=True).iloc[:40, ]
    
plt.subplots(figsize=(20,8))
sns.barplot(x=temp.IndustryVertical, y=temp.AmountInUSD)
plt.xticks(rotation=90);